Standardization: Analytical Methods for Phenolic Compounds and

main phenolic in E. purpurea roots (mean 2.27% summer, 1.68% autumn) and tops (2.02% summer,. 0.52% autumn), and echinacoside was the main phenolic ...
0 downloads 0 Views 63KB Size
1702

J. Agric. Food Chem. 2001, 49, 1702−1706

Echinacea Standardization: Analytical Methods for Phenolic Compounds and Typical Levels in Medicinal Species Nigel B. Perry,* Elaine J. Burgess, and V. leAnne Glennie Crop & Food Research, Plant Extracts Research Unit, Chemistry Department, University of Otago, P.O. Box 56, Dunedin, New Zealand

A proposed standard extraction and HPLC analysis method has been used to measure typical levels of various phenolic compounds in the medicinally used Echinacea species. Chicoric acid was the main phenolic in E. purpurea roots (mean 2.27% summer, 1.68% autumn) and tops (2.02% summer, 0.52% autumn), and echinacoside was the main phenolic in E. angustifolia (1.04%) and E. pallida roots (0.34%). Caftaric acid was the other main phenolic compound in E. purpurea roots (0.40% summer, 0.35% autumn) and tops (0.82% summer, 0.18% autumn), and cynarin was a characteristic component of E. angustifolia roots (0.12%). Enzymatic browning during extraction could reduce the measured levels of phenolic compounds by >50%. Colorimetric analyses for total phenolics correlated well with the HPLC results for E. purpurea and E. angustifolia, but the colorimetric method gave higher values. Keywords: Echinacea; phenolics; caffeic acid derivatives; high performance liquid chromatography; colorimetric analysis; medicinal herbs INTRODUCTION

Echinacea preparations, used as immunostimulants, are best-selling herbal medicines in the United States and Europe (1, 2). The U.S. National Center for Complementary and Alternative Medicine (National Institutes of Health) has identified Echinacea as an herb that warrants further testing in Phase II and/or Phase III clinical trials (3). Issues to be clarified before such trials include method of extraction and chemical standardization. Bauer (4) has suggested that Echinacea preparations should be standardized for their contents on the lipophilic alkamides and the more polar phenolic compounds. Three species of Echinacea (family Asteraceae) are generally used medicinally: E. angustifolia DC, roots (underground parts); E. pallida (Nutt.) Nutt., roots; and E. purpurea (L.) Moench, roots and tops (aerial parts) (5). The main phenolic compounds are caffeic acid derivatives (1-5, Figure 1), but there is some confusion about the different compounds found in the different medicinal species. Chicoric acid (dicaffeoyltartaric acid, also known as cichoric acid) (1) is the main phenolic in E. purpurea roots and tops, and echinacoside (2) is the main phenolic in E. angustifolia and E. pallida (6). Chicoric acid (1) has biological activities that could contribute to immunostimulatory activity, but echinacoside (2) does not seem to have any immunomodulatory relevance (4). Typical levels of 1 and 2 are given by Bauer and Wagner (6) but not levels of the minor compounds. Bauer et al. (7) pioneered the use of reversed-phase HPLC for the analysis of individual phenolic compounds in Echinacea. Other groups have used different extraction methods but similar HPLC conditions (8-12). Colorimetric analyses of “total phenolics” (13, 14) in * To whom correspondence should be addressed (telephone 64-3-479-8354; fax 64-3-479-8543; email [email protected]).

Echinacea have been used as a more rapid method for standardization, but we have not seen any published reports on this approach. In this paper we compare results from a proposed standard method from the Institute for Nutraceutical Advancement (a U.S. industry organization) for phenolics in Echinacea by HPLC (15) with a colorimetric (Folin-Denis) method, using high and low phenolics samples of E. purpurea tops and roots from a seasonal study plus E. angustifolia and E. pallida roots. We show that the ratio of chicoric acid (1) to caftaric acid (caffeoyltartaric acid, 3) differs between E. purpurea tops and roots and point out potential problems with extraction methods leading to enzymatic browning. Cynarin (4) needs to be included in the INA method for analyses of E. angustifolia roots. MATERIALS AND METHODS Plant Samples and Reference Compounds. E. purpurea tops and roots (15-18 months old) were harvested in January and April 1999 at Lincoln on the South Island of New Zealand. Four replicate plots of eight plants each were harvested at each date from a randomized block design. Tops were cut off at 10 cm above soil level, and roots comprised the rest of the plant. Samples were dried at 30 °C to moisture contents of about 10%, ground, and stored at -70 °C. Eight samples each of roots and tops were analyzed. E. angustifolia roots (18-21 months old) were harvested in May 2000 at Lincoln and Mosgiel on the South Island of New Zealand. Samples consisted of about 12 plants each. Samples were dried at 30 °C to moisture contents of about 10%, ground, and stored at -70 °C. Seven samples were analyzed. E. pallida roots (12 months old) were harvested in August 1999 at Gisborne on the North Island of New Zealand. Dried roots (moisture contents of about 10%) were stored at -70 °C. Eight individual roots were analyzed. Chlorogenic acid (5) was bought from Sigma. Echinacoside (2, impure) was donated by Prof. R. Brouillard (University Louis Pasteur, Strasbourg).

10.1021/jf001331y CCC: $20.00 © 2001 American Chemical Society Published on Web 03/03/2001

Echinacea Standardization: Phenolics Analyses

J. Agric. Food Chem., Vol. 49, No. 4, 2001 1703 20 mg injected in 4 lots, >95% pure by HPLC). The 1H NMR spectrum in CD3OD matched that reported by Horman et al. (17). Phenolics in Echinacea by the INA Method. This method follows that given on the INA Web site on May 23, 2000 (15). Extraction. Plant material was ground to pass through a 40-mesh screen (Tyler Equivalent 35 mesh or 0.0165 in.). An accurately weighed sample (ca. 0.125 g) was extracted with 25.0 mL of ethanol:water (70:30) for 15 min on an orbital shaker. The extract was then centrifuged and a 4 mL aliquot filtered through a 0.45 µm PTFE filter into an HPLC vial and capped. Standards. Chlorogenic acid (10.0 mg) was dissolved in ethanol:water (70:30) with the aid of sonication and made up to a stock concentration of 1.0 mg/mL. Further dilutions gave standard solutions for a three-point calibration curve of 0.4, 0.04, and 0.008 mg/mL. Linear regression on the HPLC analyses gave R2 values of 0.999 or better. Chromatography. A Phenomenex Prodigy column (ODS3, 5 µm, 100 Å, 4.6 × 250 mm) was used in conjunction with a Phenomenex security guard cartridge (4 × 2 mm). The column temperature was 35 °C. The mobile phases were water (containing 0.1% phosphoric acid, solvent A) and acetonitrile (Far UV Hypersolv, solvent B) in the following gradient system: initial 10% B; linear gradient to 22% B in 13 min; then to 40% B in 1 min; hold at 40% B for 0.5 min; recycle to initial conditions in 0.5 min; and hold for 5 min. The flow rate was 1.5 mL/min, injection of NaNO3 gave a tm value of 3.89 min, injection volume was 5 µL and detection was at 330 nm. The retention times of the phenolic compounds using these conditions, followed by the INA correction factors, were caftaric acid 7.5 min, 0.888; chlorogenic acid 8.2 min, 1.00; cynarin 12.2 min, 0.729 (our value); echinacoside 12.5 min, 2.22; chicoric acid 18.2 min, 0.695. Each phenolic compound was quantified using the following equation:

Figure 1. Structures of phenolic compounds in Echinacea. Chicoric acid (1) was isolated from dried ground E. purpurea roots (100 g) extracted with chloroform (1 l, 3 h) in a Soxhlet apparatus and then with methanol (1 L, 24 h). The methanol extract was evaporated in vacuo to give a yellow gum (10 g). This was dissolved in water (50 mL) containing formic acid (0.5 mL) and then extracted into ethyl acetate (3 × 300 mL) to give a gum (2 g) after evaporation of ethyl acetate. Preparative reversed phase HPLC on this (Merck LichroCart C18, 10 × 250 mm, mobile phase 4 mL/min 7:3 water: acetonitrile containing 0.1% formic acid) gave chicoric acid (1) (5.95-7.65 min, 254 nm detection, 120 mg from 400 mg injected in 20 lots). Final purification was by recrystallization from water to give chicoric acid (1) as off-white crystals (70 mg from 120 mg): mp 204-210 °C (browned and decomposed, lit. value 206 °C (16)); [R]D (c ) 0.2 g/100 mL of MeOH) -381° (lit. value -384.2° (16)); UV (MeOH) λmax 330 nm,  36 600. We could not find a value for the extinction coefficient of 1 anywhere in the literature. Analyses of a sample of 0.1 mg/ mL of chicoric acid (>95% pure by HPLC) gave an average calculated value of 0.12 mg/mL of chicoric acid (n ) 3, standard deviation (0.01 mg/mL) by the INA method. Cynarin (4) was isolated from dried ground E. angustifolia roots (200 g) extracted with chloroform (1 L, 3 h) in a Soxhlet apparatus and then with methanol (1 L, 24 h). The methanol extract was evaporated in vacuo to give a yellow gum (25 g). Lipophilic materials were removed from this by filtering through reversed phase C18 material (Sigma) eluting with water. Preparative reversed phase HPLC was then used: column Merck LichroCart C18, 10 × 250 mm; mobile phases water (containing 0.1% formic acid, solvent A) and acetonitrile (solvent B); initial 10% B, linear gradient to 25% B in 30 min, to 50% B in 5 min, recycle to initial conditions in 5 min, hold for 5 min; flow rate 5 mL/min; 330 nm detection. These conditions gave resolved peaks for echinacoside and cynarin at about 15.6 min, with cynarin (4) eluting first (1 mg from

% w/w (of dried plant material) individual phenolic compound ) (C × FV × F × 100)/W where C is the concentration of the compound (mg/mL) in the analyzed extract, calculated as chlorogenic acid from peak areas and linear regression; FV is the final volume of the analyzed extract; W is the sample weight extracted (mg); and F is the correction factor for the compound’s response against chlorogenic acid. An Excel spreadsheet was used for these calculations. A previously analyzed sample was run with each batch as an internal check. Colorimetric Analyses for Total Phenolics. Extraction. Ground plant material (1.0 g) was sonicated with ethanol: water (70:30, 80 mL) for 2 h, and then the cooled volume was adjusted to 100 mL. After the extract had settled it was centrifuged prior to spectroscopic measurements. Standards and Reagents. Chlorogenic acid (10, 30, and 50 mg) was dissolved in ethanol:water (70:30, 100 mL). The Folin-Denis reagent was prepared by mixing H2O (750 mL), Na2WO4‚2H2O (100 g), phosphomolybdic acid (H3PMo12O40, 20 g), and phosphoric acid (85%, 50 mL), heating to reflux for 2 h, cooling, and making the solution up to 1 L. The solution was stored in an amber bottle. Spectroscopy. Folin-Denis reagent (10 mL) was added to either 1 mL of extract solution or 1 mL of standard solutions. After 3 min 35% sodium carbonate solution (10 mL) was added and the test solution was made up to 100 mL with H2O and mixed. After 45 min an aliquot was centrifuged; then the clear solution was transferred into a cuvette and the absorption coefficient measured at 745 nm. The standard chlorogenic acid solutions were taken as equivalent to 1, 3, and 5% total phenolics calculated as chlorogenic acid. The total phenolics contents of the extracts were calculated using the linear regression coefficient from the standards. Statistical Analyses. Compound levels (% weight/weight) and ratios between selected compounds and between phenolics

1704 J. Agric. Food Chem., Vol. 49, No. 4, 2001

Perry et al.

Table 1. Phenolic Compound Levels in Medicinal Echinacea by the INA (15) HPLC Method and Ratio to Total Phenolic Levels by a Colorimetric (FD) Method phenolic compounda

purpurea, tops purpurea, roots lsdb (df ) 12) angustifolia, roots C.I.c (df ) 6) pallida, roots C.I.c (df ) 7)

Jan April Jan April May August

ratios

caftaric acid (3)

chlorogenic acid (5)

cynarin (4)

echinacoside (2)

chicoric acid (1)

total (1-5)

caftaric/ chicoric

totals, INA/FD

0.82 0.18 0.41 0.35 0.09